More places in DS code to bypass conda run (#9467)

For #9444
Found two more places where we could end up using conda run.
Bypass this as well to ensure we use activated environment variables instead, we know this is guaranteed to work.

This is all that's remaining as far as I know where DS code is using conda run.
We know at least from one user that conda run has failed to get list of kernels.

Author: DonJayamanne

src/client/datascience/jupyter/jupyterCommand.ts

@@ -93,7 +93,7 @@ class InterpreterJupyterCommand implements IJupyterCommand {
                     }
                 }
             }
-            return pythonExecutionFactory.createActivatedEnvironment({ interpreter: this._interpreter });
+            return pythonExecutionFactory.createActivatedEnvironment({ interpreter: this._interpreter, bypassCondaExecution: true });
         });
     }
     public interpreter(): Promise<PythonInterpreter | undefined> {

src/client/datascience/jupyter/jupyterCommandFinder.ts

@@ -368,7 +368,7 @@ export class JupyterCommandFinderImpl {
     private async createExecutionService(interpreter: PythonInterpreter): Promise<IPythonExecutionService> {
         const [currentInterpreter, pythonService] = await Promise.all([
             this.interpreterService.getActiveInterpreter(undefined),
-            this.executionFactory.createActivatedEnvironment({ resource: undefined, interpreter, allowEnvironmentFetchExceptions: true })
+            this.executionFactory.createActivatedEnvironment({ resource: undefined, interpreter, allowEnvironmentFetchExceptions: true, bypassCondaExecution: true })
         ]);
 
         // Use daemons for current interpreter, when using any other interpreter, do not use a daemon.